Skip to content

Greedy student-to-project assignment algorithm with dynamic skill weighting and tunable preference weight

Design · All designs
Evidence · 1 claim (1 for) · 1 study (1 design), q2 · 0 of 1 report an effect size · 1 claim rests on one study

Description

The matching system iteratively computes a match score for every student-project pair, combining a student's preference score weighted by an instructor-tunable parameter α with skill self-ratings weighted by a dynamic skill weight β. As the article explains, "as the algorithm progresses, and more students are assigned to the project, the value ofβp(k)for unfulfilled skills will increase leading to a higher match score for students that satisfy those skills." A skill counts as unfulfilled when no assigned student self-rated proficiency of 3 or higher on the survey's 1–5 scale. The greedy process selects the highest match score each round, with ties broken by preference then randomly, and projects with higher average preference receive more capacity up to a fixed team-size limit of five.

Design Implications

Context

Requirements

  • Instructor-defined preference weight α
  • Student skill and preference survey data
  • A predetermined skill list shared by surveys and LLM extraction

Constraints

  • The current method can favor projects with many required skills at the expense of those with rare but critical skills
  • All skills are currently treated equally, with fulfillment based on a single student's intermediate-level rating

Target Learners

  • senior-level computer science capstone students

Learning Goals

  • balanced teams with skill coverage and preference alignment

Claims

Examples

-

Key Sources

  • Brandon Pardi, Garret Castro, Michael Pisman, Avash Adhikari, and Santosh Chandrasekhar. (2026). Improving Capstone Team Outcomes through Dynamic Skill Matching and Preference Alignment. https://arxiv.org/abs/2606.15572
Page metadata
Field Value
type design
id greedy-dynamic-skill-weighting-assignment-algorithm
title Greedy student-to-project assignment algorithm with dynamic skill weighting and tunable preference weight
description The matching system iteratively computes a match score for every student-project pair, combining a student's preference score weighted by an instructor-tunable parameter α with skill self-ratings weighted by a dynamic...
status review
generated by process:settle-candidates on 2026-10-10
trust unverified — no one has confirmed this page's content
sources 1 — brandon-pardi-2026

Frontmatter is the page's machine-readable record — see the Schema & Guide.